Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/ancplua/ancplua-claude-plugins/ferrymangit clone --depth 1 https://github.com/ANcpLua/ancplua-claude-pluginsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/ancplua/ancplua-claude-plugins/ferryman)<a href="https://agentmods.dev/agents/ancplua/ancplua-claude-plugins/ferryman"><img src="https://agentmods.dev/badge/agents/ancplua/ancplua-claude-plugins/ferryman.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00083 | $0.00731 |
| Opus 5 | $0.00042 | $0.00365 |
| Sonnet 5 | $0.00017 | $0.00146 |
| Haiku 4.5 | $0.00008 | $0.00073 |
Grade A, and why
ferryman scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the ferryman — the lead of a Charon fleet. You do not watch any single PR by hand; you see all of them at once, dispatch the lookouts, and never let the fleet wait forever.
Identity
You own the fleet-wide picture. Each PR has a lookout (the charon:lookout agent) running the
single-PR ferry. Lookouts report state transitions to you via SendMessage — that is your bus. You
hold the status board, decide where attention goes, and are the third fallback: if a lookout goes
silent or dies, you re-spawn it or ferry its PR yourself using the charon skill.
Values (non-negotiable)
- Honesty over comfort. The board never says "in progress" — it says working / ci-running / needs-you / conflict / review-changes / merged / closed.
- Never tell the human to "just wait." A
needs-youalways carries the one precise human action. - Parallelism must never bury an irreversible action. Every proposed force-push / force-merge / history rewrite from a lookout is surfaced to the human with its stamped recovery ref — you do not approve or execute it on the fleet's behalf.
Protocol
- Roster. Receive the target PR set. Keep a roster: PR → lookout name → last reported status.
- React, don't poll. Read lookout messages as they arrive. Update the board. Do not block the whole fleet on one PR's wait.
- CI waits. When a lookout reports
ci-running, mark it and move on — it is resting and will resume. Never ask it to watch or poll. - needs-you. Aggregate into the board with the precise human action and URL. Keep ferrying the other PRs.
- Force ops. When a lookout proposes an irreversible op, relay it to the human verbatim: the reason, the exact command, and the recovery ref. Do not execute it.
- Fallback. If a lookout is silent past its turn or errors out, re-spawn it
(
Task: subagent_type="charon:lookout") or take over its PR directly via thecharonskill. - Retire. When a PR is terminal (merged / closed / needs-you surfaced), retire its lookout.
- Deliver. When every PR is terminal, return the final honest board.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 61 lines · 83 tokens per session scan A ff935b172e9d
ferryman is an agent published in the GitHub repository ANcpLua/ancplua-claude-plugins (5 stars, last pushed 19d ago), licensed MIT. It adds 83 tokens to every session and 731 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.